The topic of the thesis is optimization of distributed intrusion detection system (IDS) in wireless sensor network (WSN) by use of evolution metaheuristics. The thesis should include an overview of modern evolutionary algorithms, with focus on multi-criteria optimization (e.g., algorithms NSGA-II [1], SPEA2 [2]). In addition it should contain implementation of a system that should enable effective optimization of IDS for WSNs by combining ParadisEO [3] (framework for evolutionary metaheuristics) and Boinc [4] (system for distributed computation). The IDS for WSN should be simulated and evaluated by simulator MiXiM [5]. The implemented system should be easily configurable for various optimization scenarios. It should allow the user to configure parameters to be optimized and define optimization objectives. The system should be tested on a simple scenario (e.g., laboratory testbed) using algorithms NSGA-II and SPEA2. The algorithms should be compared with emphasis on speed of convergence and diversification of the found solutions.

[1] http://www.iitk.ac.in/kangal/Deb_NSGA-II.pdf
[2] http://www.kddresearch.org/Courses/Spring-2007/CIS830/Handouts/P8.pdf
[3] http://paradiseo.gforge.inria.fr/
[4] http://boinc.berkeley.edu/
[5] http://mixim.sourceforge.net/